ClawHub
Back to skill

Security audit

CleanShot Skill

Security checks for vulnerabilities and agentic risk

Overview

This is an instruction-only CleanShot workflow skill whose screen capture, OCR, and recording guidance is disclosed and user-directed.

Install only if you are comfortable granting the separate CleanShot Tool plugin access to capture or record visible screen content. Review the screen before sharing captures, and use CleanShot Cloud upload only when you explicitly want a shareable link.

A.I.G

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (3)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The examples repeatedly instruct screenshot, OCR, and screen-recording actions but never warn that these operations may capture secrets such as passwords, tokens, private chats, or other sensitive on-screen data. In a workflow-guidance skill, this omission can normalize broad capture behavior and increase the chance that users or downstream agents expose confidential information unintentionally.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
76% confidence
Finding

The changelog states that the skill guidance was rewritten as "English-first" instructions and later notes bilingual examples while keeping core instructions in English. This suggests a language policy preference that may constrain user interaction without explicitly indicating user choice or opt-in.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
86% confidence
Finding

The phrase 'These examples are English-first' establishes a preferred language behavior in natural language. While multilingual examples are later included, this line still expresses a default language policy without explaining why that constraint is necessary or making it explicitly user-selectable.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.